Building financial projections that aren't fiction

Every projection starts at the top line, and the top line is where most forecasts already go wrong. The lazy way is to type 'revenue grows 15% a year' and move on. The honest way is to build revenue from its underlying drivers — the real, physical things that produce sales — so that every rupee of forecast revenue has a reason behind it.

The fundamental driver: volume × price

At its core, almost every business's revenue is volume × price — how many units it sells, times what it charges. Forecasting each separately is far more honest than forecasting the product, because volume and price have different stories. Volume grows with demand, capacity and distribution; price grows with inflation, premiumisation and pricing power. Split them and you can sanity-check each.

Other lenses on the same question

• Cohorts / subscriptions. For a subscription or app business, build revenue from customers: how many you start with, how many you add, how many churn, and what each pays. Revenue = active customers × average revenue per customer.
• Capacity-led. For a factory, hotel or airline, revenue is capped by capacity × utilisation × price (rooms × occupancy × rate). You can't forecast more output than the asset can physically produce.
• Same-store + new-store. For retail chains, split growth into existing stores (same-store sales growth) and new stores opened — two very different drivers that lazy 'X% growth' hides.